[Buildroot] [PATCH] dockstar bsp

[Thomas Petazzoni <thomas.petazzoni@free-electrons.com>]

Dear Steve Calfee,

On Sat, 29 Dec 2012 14:45:39 -0800, Steve Calfee wrote:
> Simple bsp for the Seagate Dockstar.
> 
> Builds and runs using buildroot 2012.11
> 
> Signed-off-by: Steve Calfee <stevecalfee@gmail.com>
> ---
>  board/sheevaplug/Config.in                         |   13 +
>  board/sheevaplug/Makefile.in                       |    4 +
>  board/sheevaplug/seagatedockstar/Makefile.in       |    5 +

No, we really don't want per-board configuration options or Makefiles.
A board is just a particular configuration of Buildroot.

Moreover your Makefile.in files still reference some old files in
target/device/ that no longer exist.

>  board/sheevaplug/seagatedockstar/README            |   62 +++++

The README file should not be a story of how the thing was written, but
simply how to use this configuration on this particular hardware (how
to flash the images, etc.).

>  .../dockstar_dev_external_toolchain_defconfig      |   24 ++
>  .../dockstar_internal_toolchain_defconfig          |   21 ++
>  .../dockstar_toolchain_only_defconfig              |   16 ++

Those configurations are useless as they are not visible through "make
help" or "make <foobar>_defconfig". And anyway, we only want one
configuration per board.

>  .../seagatedockstar/linux-3.2plus.config           |  294 ++++++++++++++++++++
>  board/sheevaplug/seagatedockstar/patchrootfs.sh    |   14 +
>  .../seagatedockstar/skeleton/etc/inittab           |   46 +++
>  .../sheevaplug/seagatedockstar/skeleton/etc/shadow |   12 +

Why is it needed to override the default inittab and shadow?

>  configs/dockstart_defconfig                        |   21 ++

> --- /dev/null
> +++ b/configs/dockstart_defconfig
> @@ -0,0 +1,21 @@
> +BR2_arm=y
> +BR2_arm926t=y

Ok for those two ones.

> +BR2_JLEVEL=4
> +BR2_PACKAGE_GDB=y
> +BR2_ENABLE_LOCALE_PURGE=y
> +BR2_TOOLCHAIN_BUILDROOT_LARGEFILE=y
> +BR2_TOOLCHAIN_BUILDROOT_INET_IPV6=y
> +BR2_TOOLCHAIN_BUILDROOT_INET_RPC=y
> +BR2_TOOLCHAIN_BUILDROOT_LOCALE=y
> +BR2_TOOLCHAIN_BUILDROOT_CXX=y
> +BR2_TARGET_GENERIC_HOSTNAME="dockstar"
> +BR2_TARGET_GENERIC_ISSUE="Welcome to Dockstar and Buildroot"
> +BR2_ROOTFS_POST_BUILD_SCRIPT="$(TOPDIR)/board/sheevaplug/seagatedockstar/patchrootfs.sh"

I don't think any of these is needed.

> +BR2_TARGET_ROOTFS_UBIFS=y
> +BR2_TARGET_ROOTFS_UBIFS_RT_ZLIB=y
> +BR2_TARGET_ROOTFS_UBI=y
> +BR2_TARGET_ROOTFS_TAR_GZIP=y
> +BR2_LINUX_KERNEL=y
> +BR2_LINUX_KERNEL_USE_CUSTOM_CONFIG=y
> +BR2_LINUX_KERNEL_CUSTOM_CONFIG_FILE="$(topdir)board/sheevaplug/seagatedockstar/linux-3.2plus.config"
> +BR2_LINUX_KERNEL_INSTALL_TARGET=y

Ok. Is installation of the kernel image inside the target filesystem
really needed? How does this board boots?

To summarize, a board configuration in Buildroot is just:

 * A minimal defconfig in configs/<blabla>_defconfig, that selects the
   proper architecture, proper bootloader, filesystem image format and
   kernel version/configuration. No toolchain options, no packages, etc.

 * Possibly in board/<manufacturer>/<boardname>/, a kernel
   configuration file, a README file, and other files that have to be
   added in the root filesystem to make the board work.

Thanks!

Thomas
-- 
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com